ABSTRACT
In this chapter, the author shows how to model data with more than one categorical predictor. Sometimes the data can arise from observational studies but such data more commonly arises from designed experiments, often called factorial designs. If all possible combinations of the levels of the factors occur at least once. Replication can be expensive, so sometimes it is better to use the experimental resources to investigate more factors. This leads us to an example with many factors but no replication. Mazumdar and Hoa report an experiment to test the strength of a thermoplastic composite depending on the power of a laser and the speed of a tape. Fractional factorials use only a fraction of the number of runs in a full factorial experiment. This is done to save the cost of the full experiment or to make only a few runs because the experimental material is limited.
